Social network analysis

Pablo Barberá
January 22, 2016

Moreno, Who Shall Survive?, 1934

Christakis & Fowler, NEJM, 2007

Adamic & Glance, 2004, IWLD

Email network of a company

Barberá et al, 2015, Psychological Science

Outline

Social network analysis:

Basic concepts
Importing network data into R
Network visualization
Node and network-level summary statistics
Community detection
Network modeling

Basic concepts

Node (vertex): each of the objects in the graph

Edge (tie): each of the connections between nodes.

Two types of edges:

Undirected: symmetric connection, represented by lines.
Directed: imply direction, represented by arrows.

A graph consists of a set of nodes and edges.

A few examples

Classroom: students / friendships
Twitter: users / retweets
Academic literature: papers / citations
Internet: websites / hyperlinks
Trade: countries / trade flows
Biology: neurons / connections

Network visualization:

plot of chunk unnamed-chunk-1

Adjacency matrix:

         Jennifer Pablo Dan Kevin Denis
Jennifer        0     1   1     0     0
Pablo           1     0   0     1     1
Dan             1     0   0     1     0
Kevin           0     1   1     0     1
Denis           0     1   0     1     0

Edgelist:

     Node1      Node2  
[1,] "Jennifer" "Pablo"
[2,] "Jennifer" "Dan"  
[3,] "Pablo"    "Kevin"
[4,] "Pablo"    "Denis"
[5,] "Kevin"    "Denis"
[6,] "Dan"      "Kevin"